Professional Documents
Culture Documents
WP 160620 Basic in Tror He Ology
WP 160620 Basic in Tror He Ology
Viscosity
There are two basic types of flow, these being shear flow and extensional flow.
In shear flow fluid components shear past one another while in extensional flow
fluid component flowing away or towards from one other. The most common
flow behavior and one that is most easily measured on a rotational rheometer
or
viscometer is shear flow and this viscosity introduction will focus on this behavior
and how to measure it.
Shear Flow
Shear flow can be depicted as layers of fluid sliding over one another with each
layer moving faster than the one beneath it. The uppermost layer has maximum
velocity while the bottom layer is stationary. For shear flow to take place a shear
force must act on the fluid. This external force takes the form of a shear stress (σ)
which is defined as the force (F) acting over a unit area (A) as shown in Figure 1.
In response to this force the upper layer will move a given distance x, while the
bottom layer remains stationary. Hence we have a displacement gradient across
the sample (x/h) termed the shear strain (γ). For a solid which behaves like a single
block of material, the strain will be finite for an applied stress – no flow is possible.
However, for a fluid where the constituent components can move relative to one
another, the shear strain will continue to increase for the period of applied
stress.
This creates a velocity gradient termed the shear rate or strain rate ( ) which is
the rate of change of strain with time (dγ/dt).
Figure 1 – Quantification of shear rate and shear stress for layers of fluid sliding over one another
Newtonian fluids are fluids in which the shear stress is linearly related to the
shear rate and hence the viscosity is invariable with shear rate or shear stress.
Typical Newtonian fluids include water, simple hydrocarbons and dilute colloidal
dispersions. Non-Newtonian fluids are those where the viscosity varies as a
function of the applied shear rate or shear stress. It should be noted that fluid
viscosity is both pressure and temperature dependent, with viscosity generally
increasing with increased pressure and decreasing temperature. Temperature
is more critical than pressure in this regard with higher viscosity fluids such as
asphalt or bitumen much more temperature dependent than low viscosity fluids
such as water.
2 A Basic Introduction to
WHITEPAPE
capable of working in controlled stress or controlled rate mode which means it is
3 A Basic Introduction to
WHITEPAPE
possible to apply a torque and measure the rotational speed or alternatively apply
a rotational speed and measure the torque required to maintain that speed. In
controlled stress mode a torque is requested from the motor which translates to
a force (F) acting over the surface area of the plate (A) to give a shear stress (F/A).
In response to an applied shear stress a liquid like sample will flow with a shear
rate dependent on its viscosity. If the measurement gap (h) is accurately known
then the shear rate (V/h) can be determined from the measured angular velocity
(ω) of the upper plate, which is determined by high precision position sensors,
and its radius (r), since V = r ω. Other measuring systems including cone-plate
and concentric cylinders are commonly used for measuring viscosity with cone-
plate often preferred since shear rate is constant across the sample. The type of
measuring system used and its dimensions is dependent on the sample type and
its viscosity. For example, when working with large particle suspensions a cone-
plate system is often not suitable.
Figure 2 – Illustration showing a sample loaded between parallel plates and shear profile
generated across the gap
Shear thinning
The most common type of non-Newtonian behavior is shear thinning or
pseudoplastic flow, in which the fluid viscosity decreases with increasing shear.
At low enough shear rates, shear thinning fluids will show a constant viscosity
value, η0, termed the zero shear viscosity or zero shear viscosity plateau. At a
critical shear rate or shear stress, a large drop in viscosity is observed, which
signifies the beginning of the shear thinning region. This shear thinning region
can be mathematically described by a power law relationship which appears as a
linear section when viewed on a double logarithmic scale (Figure 5), which is how
rheological flow curves are often presented. At very high shear rates a second
constant viscosity plateau is observed, called the infinite shear viscosity plateau.
This is given the symbol η∞ and can be several orders of magnitude lower than η0
depending on the degree of shear thinning.
Some highly shear-thinning fluids also appear to have what is termed a yield
stress, where below some critical stress the viscosity becomes infinite and hence
characteristic of a solid. This type of flow response is known as plastic flow and is
characterized by an ever increasing viscosity as the shear rate approaches zero
(no visible plateau). Many prefer the description ‘apparent yield stress’ since some
materials which appear to demonstrate yield stress behavior over a limited shear
rate range may show a viscosity plateau at very low shear rates.
4 A Basic Introduction to
WHITEPAPE
Figure 3 - Typical flow curves for shear thinning fluids with a zero shear viscosity and an apparent
yield stress
Why does shear thinning occur? Shear thinning is the result of micro-structural
rearrangements occurring in the plane of applied shear and is commonly
observed for dispersions, including emulsions and suspensions, as well as
polymer solutions and melts. An illustration of the types of shear induced
orientation which can occur for various shear thinning materials is shown in
Figure 4.
Figure 4 - Illustration showing how different microstructures might respond to the application of
shear
At low shear rates materials tend to maintain an irregular order with a high
zero shear viscosity (η0) resulting from particle/molecular interactions and the
restorative effects of Brownian motion. In the case of yield stress materials such
interactions result in network formation or jamming of dispersed elements
which must be broken or unjammed for the material to flow. At shear rates
or stresses high enough to overcome these effects, particles can rearrange or
reorganize in to string-like layers, polymers can stretch out and align with the
flow, aggregated structures can be broken down and droplets deformed from
their spherical shape. A consequence of these rearrangements is a decrease in
molecular/particle interaction and an increase in free space between dispersed
components, which both contribute to the large drop in viscosity. η∞ is associated
with the maximum degree of orientation achievable and hence the minimum
5 A Basic Introduction to
WHITEPAPE
attainable viscosity and is influenced largely by the solvent viscosity and related
hydrodynamic forces.
Model fitting
The features of the flow curves shown in Figure 3 can be adequately modeled
using some relatively straight forward equations. The benefits of such an
approach are that it is possible to describe the shape and curvature of a flow
curve through a relatively small number of fitting parameters and to predict
behavior at unmeasured shear rates (although caution is needed when using
extrapolated data). Three of the most common models for fitting flow curves
are the Cross, Power law and Sisko models. The most applicable model largely
depends on the range of the measured data or the region of the curve you would
like to model (Figure 5). There are a number of other models available such as the
Carreau-Yasuda model and Ellis models for example. Other models accommodate
the presence of a yield stress, these include Casson, Bingham, and Herschel-
Bulkley models.
η0 is the zero shear viscosity; η∞ is the infinite shear viscosity; K is the cross
constant, which is indicative of the onset of shear thinning; m is the shear
thinning index, which ranges from 0 (Newtonian) to 1 (Infinitely shear thinning);
n is the power law index which is equal to (1 – m), and similarly related to the
extent of shear thinning, but with n → 1 indicating a more Newtonian response;
k is the consistency index which is numerically equal to the viscosity at 1 s-1.
Figure 5 – Illustration of a flow curve and the relevant models for describing its shape
6 A Basic Introduction to
WHITEPAPE
Shear thickening
While most suspensions and polymer structured materials are shear thinning,
some materials can also show shear thickening behavior where viscosity increases
with increasing shear rate or shear stress. This phenomenon is often called
dilatancy, and although this refers to a specific mechanism for shear thickening
associated with a volume increase, the terms are often used interchangeably.
In most cases, shear thickening occurs over a decade of shear rates and there
can be a region of shear thinning at lower and higher shear rates. Usually
dispersions or particulate suspensions with high concentration of solid particles
exhibit
shear thickening. Materials exhibiting shear thickening are much less common in
industrial applications than shear thinning materials. They do have some useful
applications such as in shock absorbers and high impact protective equipment
but for the most part shear thickening is an unwanted effect which can lead to
major processing issues.
For suspensions, shear thickening generally occurs in materials that show shear
thinning at lower shear rates and stresses. At a critical shear stress or shear rate
the organized flow regime responsible for shear thinning is disrupted and so
called ‘hydro-cluster’ formation or ‘jamming’ can occur. This gives a transient
solid-like response and an increase in the observed viscosity. Shear thickening
can also occur in polymers, in particular amphiphilic polymers, which at high shear
rates may open-up and stretch, exposing parts of the chain capable of forming
transient intermolecular associations.
Thixotropy
For most liquids shear thinning is reversible and the liquids will eventually gain
their original viscosity when the shearing force is removed. When this recovery
process is sufficiently time dependent the fluid is considered to be thixotropic.
Thixotropy is related to the time dependent microstructural rearrangements
occurring in a shear thinning fluid following a step change in applied shear (Figure
6). A shear thinning material may be thixotropic but a thixotropic material will
always be shear thinning. A good practical example of a thixotropic material is
paint. A paint should be thick in the can when stored for long periods to prevent
separation, but should thin down easily when stirred for a period time – hence it
is shear thinning. Most often its structure does not rebuild instantaneously on
ceasing stirring – it takes time for the structure and hence viscosity to rebuild to
give sufficient working time.
7 A Basic Introduction to
WHITEPAPE
The best way to evaluate and quantity thixotropy is using a three step shear test
as shown in Figure 7. A low shear rate is employed in stage one which is meant
to replicate the samples at near rest behavior. In stage two a high shear rate is
applied for a given time to replicate the breakdown of the sample's structure
and can be matched to the process of interest. In the third stage the shear rate
is again dropped to a value generally equivalent to that employed in stage one
and viscosity recovery followed as a function of time. To compare thixotropic
behavior between samples the time required to recover 90% (or a defined
amount) of
the initial viscosity can be used. This time can therefore be viewed as a relative
measure of thixotropy - a small rebuild time indicates that the sample is less
thixotropic than a sample with a long rebuild time.
Figure 7 - Illustration showing a step shear rate test for evaluating thixotropy and expected
response for non-thixotropic and thixotropic fluids
Yield Stress
Many shear thinning fluids can be considered to possess both liquid and solid
like properties. At rest these fluids are able to form intermolecular or
interparticle networks as a result of polymer entanglements, particle association,
or some other interaction. The presence of a network structure gives the
material predominantly solid like characteristics associated with elasticity, the
strength
of which is directly related to the intermolecular or interparticle forces (binding
force) holding the network together, which is associated with the yield stress.
If an external stress is applied which is less than the yield stress the material will
deform elastically. However, when the external stress exceeds the yield stress
the network structure will collapse and the material will begin to flow as if it is
8 A Basic Introduction to
WHITEPAPE
a liquid. Despite yield stress clearly being apparent in a range of daily activities
9 A Basic Introduction to
WHITEPAPE
Figure 8 – Shear stress/shear rate plots depicting various types of flow behavior
There are a number of experimental tests for determining yield stress, including
multiple creep testing, oscillation amplitude sweep testing and also steady shear
testing; the latter usually with the application of appropriate models such as the
Bingham, Casson and Herschel-Bulkley models.
Where σY is the yield stress and ηB the Bingham viscosity, represented by the
slope of shear stress versus shear rate in the Newtonian region, post yield. The
Herschel-Bulkley model is just a power law model with a yield stress term and
hence represents shear thinning post yield, with K the consistency and n the
power law index. All of the various tests for measuring yield stress are discussed
in [5].
One of the quickest and easiest methods for measuring the yield stress is to
perform a shear stress ramp and determine the stress at which a viscosity peak
is observed (Figure 9). Prior to this viscosity peak the material is undergoing
elastic deformation where the sample is simply stretching. The peak in viscosity
1 A Basic Introduction to
WHITEPAPE
represents the point at which this elastic structure breaks down (yields) and the
material starts to flow. If there is no peak this indicates that the material does
not have a yield stress under the conditions of the test.
Yield stress can be related to the stand-up properties (slump) of a material, the
stability of a suspension, or sagging of a film on a vertical surface, as well as many
other applications.
Figure 9 – Linear shear stress ramp and shear strain response (left) and corresponding viscosity
against shear stress for materials with and without a yield stress
Viscoelasticity
As the name suggests, viscoelastic behavior describes materials which show
behavior somewhere between that of an ideal liquid (viscous) and ideal solid
(elastic). There are a number of rheological techniques for probing the viscoelastic
behavior of materials, including creep testing, stress relaxation and oscillatory
testing. Since oscillatory shear rheometry is the primary technique that is used
to measure viscoelasticity on a rotational rheometer this will be discussed in
greatest detail, although creep testing will be also introduced.
Elastic behaviour
Structured fluids have a minimum (equilibrium) energy state associated with
their ‘at rest’ microstructure. This state may relate to inter-entangled chains in
a polymer solution, randomly ordered particles in a suspension, or jammed
droplets in an emulsion. Applying a force or deformation to a structured fluid
will shift the equilibrium away from this minimum energy state, creating an
elastic force that tries to restore the microstructure to its initial state. This is
analogous to a stretched spring trying to return to its undeformed state.
Figure 10 – The response of an ideal solid (spring) to the application and subsequent removal of a
strain inducing force
1 A Basic Introduction to
WHITEPAPE
A spring is representative of a linear elastic solid that obeys Hooke’s law, in that
the applied stress is proportional to the resultant strain as long as the elastic
limit is not exceeded, and will return to its initial shape when the stress is removed
as shown in Figure 10. If the elastic limit is surpassed the spring will be
permanently distorted. These same principles can also be applied to simple shear
deformation as illustrated in Figure 11.
Figure 11 – Quantification of stress, and strain for an ideal solid deforming elastically in shear
Viscous Behaviour
Just as a spring is considered representative of a linear elastic solid that obeys
Hooke’s law, a viscous material can be modeled using a dashpot which obeys
Newton’s law. A dashpot is mechanical device consisting of a plunger moving
through a viscous Newtonian fluid.
Figure 12 – Response of an ideal liquid (dashpot) to the application and subsequent removal of
a strain inducing force
When a stress (or force) is applied to a dashpot, the dashpot immediately starts
to deform and goes on deforming at a constant rate (strain rate) until the
stress is removed (Figure 12). The energy required for deformation or
displacement is
1 A Basic Introduction to
WHITEPAPE
dissipated within the fluid (usually as heat) and the strain is permanent. The strain
evolution in an ideal liquid is given by the following expression:
Viscoelastic Behaviour
A vast majority of materials show rheological behavior that classifies them in a
region somewhere between that of liquids and solids and are therefore classed
as viscoelastic materials. Consequently it is possible to combine springs and
dashpots in such a way as to model or describe real viscoelastic behavior. The
simplest representation of a viscoelastic liquid is a spring and dashpot connected
in series, which is called the Maxwell model. A viscoelastic solid can be similarly
represented by the Kelvin-Voigt model, which utilizes the same combination of
elements but connected in parallel (Figure 13).
Figure 13 – (left) Maxwell model representative of a simple viscoelastic liquid; (right) Kelvin-Voigt
model representative of a simple viscoelastic solid
If a stress is applied to a Maxwell model then at very short times the response
is predominantly elastic and governed by G, while at much longer times viscous
behavior prevails and is governed by η. The strain evolution in a Maxwell model
can be described by the following expression.
1 A Basic Introduction to
WHITEPAPE
The model which best describes the viscoelastic behavior of real systems in
response to an applied stress is the Burgers model (Figure 14), which is essentially
a Maxwell and Kelvin-Voigt model connected in series.
Creep Testing
The test protocol described in the previous section whereby a constant stress
is applied to a viscoelastic material and the strain response measured is what
is called a creep test. This kind of measurement is usually applied to solid-like
materials like metals, which creep on long timescales rather than flow, although
the test is applicable to all kinds of viscoelastic material. The test involves applying
a constant shear stress over a period of time and measuring the resultant shear
strain. The test must be performed in the linear viscoelastic region (see next
section) where the microstructure remains intact. The measured response in a
creep test is usually presented in terms of the creep compliance J(t) which is the
ratio of the measured strain to the applied stress, or inverse modulus.
A typical creep and recovery profile for a material showing Burgers type behavior
is shown in Figure 15. An initial elastic response is first observed, followed by a
delayed elastic response and finally a steady state (linear) viscous response at
1 A Basic Introduction to
WHITEPAPE
longer times. The gradient of this line is equal to the strain rate and can
therefore be used to calculate the zero shear viscosity of the fluid. If the steady
state
linear response is extrapolated back to zero time then the intercept is equal
to the equilibrium compliance (JE). This is the compliance or strain response
associated with just the elastic components of the material i.e. springs in the
Burgers model. The recovery step begins once steady state has been attained
and involves removing the applied stress and monitoring the strain as the
stored elastic stresses relax. Only the elastic deformation of the sample is able to
recover fully because the viscous deformation is permanent and JR the recovery
compliance should eventually equal JE . To accurately model the response of
real systems in creep testing it is often necessary to use multiple Kelvin-Voigt
elements.
Figure 15 – (right) representation of a Burger model and (left) expected profile of a Burger model
undergoing creep and recovery testing with equilibrium compliance (J) and recovery compliance
(J)
Figure 16 – Expected creep response for a viscoelastic liquid and viscoelastic solid
1 A Basic Introduction to
WHITEPAPE
Figure 17 – Illustration showing a sample loaded between parallel plates with an oscillatory
(sinusoidal) shear profile applied
To perform oscillation testing with a parallel plate measuring system, the sample
is loaded between the plates at a known gap (h) and the upper plate oscillated
back and forth at a given stress or strain amplitude and frequency (Figure
17). This motion can be represented as a sinusoidal wave with the stress or
strain amplitude plotted on on the y-axis and time on the x-axis. In a
controlled stress measurement an oscillating torque is applied to the upper
plate and the resultant angular displacement measured, from which the strain
is calculated. In a controlled strain experiment the angular displacement is
controlled and the torque required to give that displacement is measured,
from which the shear stress can be calculated.
The ratio of the applied stress (or strain) to the measured strain (or stress) gives
the complex modulus (G*), which is a quantitative measure of material stiffness or
resistance to deformation, where
For a purely elastic material (stress is proportional to strain) the maximum stress
occurs at maximum strain (when deformation is greatest) and both stress and
strain are said to be in phase. For a purely viscous material (stress is proportional
to strain rate) the maximum stress occurs when the strain rate is maximum (when
flow rate is greatest) and stress and strain are out of phase by 90° or π/2 radians
(quarter of a cycle). For a viscoelastic material the phase difference between
stress and strain will fall somewhere between the two extremes. This is illustrated
in Figure 18.
1 A Basic Introduction to
WHITEPAPE
Figure 18 – Stress and strain wave relationships for a purely elastic (ideal solid), purely viscous
(ideal liquid) and a viscoelastic material
It is this phase difference which allows the viscous and elastic components
contributing to the total material stiffness (G*) to be determined; the phase
angle δ being a relative measure of the materials viscous and elastic
characteristics. For a purely elastic material, δ will have a value equal to 0°, while
a purely viscous material will have a δ value equal to 90°. Viscoelastic materials
having both characteristics will have a δ value between 0 and 90°, with 45°
representing the boundary between solid-like and liquid-like behavior. This value
may be considered indicative of a gel (or sol) point, which signifies the onset of
network formation (or breakdown). Phase angle is often expressed as the loss
tangent (tan δ) particularly when working with polymer systems.
1 A Basic Introduction to
WHITEPAPE
Figure 20 – Argand diagram showing the relationship between G’ and G” and G* in the complex
plane
As with G* this can be broken down into its component parts, which include the
dynamic viscosity (η’) and the storage viscosity (η”), which represent the real and
imaginary parts of η* respectively.
1 A Basic Introduction to
WHITEPAPE
Figure 21 – Illustration showing the LVER for different materials as a function of applied strain
1 A Basic Introduction to
WHITEPAPE
Figure 22 – Typical frequency response for a viscoelastic solid, viscoelastic liquid and a gel in
oscillatory testing
For a gel like material G’ and G” are parallel and δ is constant with a value
between 0 and 45º. A suitable mechanical model for describing gel-like behavior
is a spring in parallel with a Maxwell element. Both viscoelastic solid and gel-like
systems show yield stress behavior since they require any associated structure
(represented by single springs in their respective models) to be broken for
macroscopic flow to occur.
2 A Basic Introduction to
WHITEPAPE
Another technique which can be used for reducing the time required for collection
of frequency data is creep testing. Although not mathematically straightforward
there are algorithms for transforming J(t) to G’(ω) and G”(ω) and their associated
parameters. Microrheological techniques can also be used for extending
viscoelastic measurements to high frequencies especially for weakly structured
materials.
Another important rule which allows steady shear viscosity data to be predicted
from oscillatory data is the Cox-Merz rule, which states that the complex viscosity
as a function of frequency is equivalent to the steady shear viscosity as a function
of shear rate. This rule seems to hold for simple solutions including polymer
melts, however, more complex dispersions may show variations.
References
3. Larson RG; The Structure and Rheology of Complex Fluids, Oxford University
Press, New York (1999)
9. Duffy JJ, Rega CA, Jack R, Amin S; An algebraic approach for determining
viscoelastic moduli from creep compliance through application of the Generalised
Stokes-Einstein relation and Burgers model, Appl. Rheol. 26:1 (2016)
2 A Basic Introduction to
WHITEPAPE
Malvern Instruments Malvern Instruments is part of Spectris plc, the Precision Instrumentation and Controls Company.
Limited
Grovewood Road,
Spectris and the Spectris logo are Trade Marks of Spectris plc.
Malvern, Worcestershire,
UK. WR14 1XZ
Malvern Instruments pursues a policy of continual improvement due to technical development. We therefore
reserve the right to deviate from information, descriptions, and specifications in this publication without notice.
Malvern Instruments shall not be liable for errors contained herein or for incidental or consequential damages in
connection with the furnishing, performance or use of this material.
Malvern Instruments owns the following registered trademarks: Bohlin, FIPA, Insitec, ISYS, Kinexus, Malvern,
Malvern 'Hills' logo, Mastersizer, MicroCal, Morphologi, Rosand, 'SEC-MALS', Viscosizer, Viscotek, Viscogel and
Zetasizer.